CHO, mainly muscle glycogen, is the predominant substrate used during prolonged exercise at intensities greater than 65% of maximal oxygen uptake (V˙O2max) (36,43,44), and glycogen depletion is closely linked with the development of fatigue. Consuming a CHO-rich diet can increase muscle glycogen content (3,14,23), delay the onset of fatigue (3), and maintain performance during periods of intensive training (1). Therefore, endurance athletes consume much CHO in the belief that training longer and/or more intensely, as a result of increased glycogen availability, will promote superior training adaptations (15).
Regular endurance training induces several adaptations that enhance performance, including an increase in V˙O2max and a shift in substrate metabolism from CHO to lipid oxidation (17,27). Such changes are largely the result of increased mitochondrial volume (20), increased capillary density (27), and greater activity of oxidative enzymes such as citrate synthase and β-hydroxyacyl-CoA dehydrogenase (β-HAD) (11,27). These adaptations occur as a result of an accumulation of specific proteins, and thus, altered gene expression is considered to be an important process underlying the adaptive response to endurance exercise (12). Interestingly, it seems that altered substrate availability might play a role in the regulation of exercise-induced gene expression. For example, commencing an acute bout of endurance exercise with low muscle glycogen induces a greater increase in the transcriptional activity of several metabolic genes when compared with exercise performed with normal glycogen stores (10,24,33).
Furthermore, Hansen et al. (12) reported that untrained subjects who completed 10 wk of aerobic training, 50% of which commenced with low muscle glycogen, had more pronounced increases in resting muscle glycogen content and citrate synthase activity when compared with the same volume of training performed with normal glycogen concentrations. Training with low muscle glycogen also resulted in a twofold greater increase in exercise capacity (time to fatigue at 90% peak power output) when compared with the same volume of training undertaken with normal glycogen. These findings demonstrate that, under certain circumstances, training with low muscle glycogen can be beneficial. However, some details make it difficult to extrapolate these findings. First, the subjects were untrained, and it is not yet known if training with low muscle glycogen will translate into improved adaptations in already well-trained athletes. Second, the subjects performed a fixed amount of work in each training bout, although higher glycogen stores would allow more exercise and/or higher intensities. Third, the results are from single-leg kicking exercise and not a competitive sport. Considering these points, further research is clearly needed before glycogen depleted training can be recommended as a practical strategy to enhance training adaptations and performance in already well-trained athletes.
One recent study (49), using a similar training model to that of the study of Hansen et al. (i.e., training twice every second day vs once daily), but in well-trained cyclists, found that self-selected training intensity was reduced when high-intensity interval training (HIT) commenced with low muscle glycogen. However, after completing the training period, time trial cycling performance was improved to a similar extent in high and low glycogen groups. Another interesting observation was that training with low muscle glycogen resulted in higher rates of whole-body fat oxidation during exercise, whereas training with high muscle glycogen had no effect on substrate metabolism. The aim of the present study was twofold. First, we aimed to confirm the recent findings of Yeo et al. (49). Second, using stable isotope tracers, we aimed to examine in greater detail the reported changes in substrate metabolism after training with low muscle glycogen. Specifically, by infusing [U-13C]palmitate and [6,6-2H2]glucose, we were able to quantify plasma and muscle-derived CHO and fat oxidation and thus gain a better understanding of adaptations in substrate metabolism after training with low muscle glycogen.
METHODS
Subjects
Fourteen endurance-trained male cyclists who had not undertaken any high-intensity interval training in the 4 wk before the investigation were recruited to participate in this study. Subjects' characteristics can be seen in Table 1. Subjects were informed of the potential risks involved with the experimental procedures before providing their written consent. The study was approved by the local research ethics committee (Solihull REC, UK).
;)
TABLE 1: Subjects' characteristics.
Preliminary Testing
Before baseline measurements, subjects visited the laboratory on two separate occasions. During the first visit, subjects performed an incremental exercise test to exhaustion on an electromagnetically braked cycle ergometer (Lode Excalibur Sport, Groningen, The Netherlands) for determination of V˙O2max and maximum power output (Wmax). The test began at 95 W, followed by 35-W increments every 3 min thereafter. Gas exchange measurements were performed during the final minute of each stage using traditional Douglas bag techniques, with V˙O2 and V˙CO2 calculated using conventional equations. Approximately 3-5 d later, subjects returned to the laboratory, where they were infused with [1,2-13C]acetate during the 60-min rest and 60-min cycling at 70% V˙O2max. Expired breath samples were collected every 15 min for measures of V˙O2, V˙CO2, and 13CO2 enrichment. The acetate infusion was performed to determine an individual acetate recovery factor for each subject to correct [U-13C]palmitate oxidation rates for carbon label retention within the bicarbonate pools and by way of isotopic exchange reactions within the tricarboxylic acid cycle (acetate recovery was, on average, 89%). After 2 h of infusion, and on completion of the cycle task, the acetate infusion was stopped, and subjects began a 60-min time trial. Subjects were later pair-matched for V˙O2max, Wmax, time trial performance, and training history and were randomly assigned to either HIGH- or LOW-glycogen training groups.
General Experimental Design
The experiment consisted of three separate phases: 1) baseline measures, 2) a 3-wk training intervention, and 3) posttraining measures.
Baseline measures.
Muscle biopsy.
A muscle biopsy (∼80 mg) was obtained from the vastus lateralis under local anesthetic (2% lidocaine) using the percutaneous needle biopsy technique (2) modified for use with suction (9). Samples were immediately frozen in liquid nitrogen and stored at −80°C until later analysis. All muscle biopsies were obtained in the morning after an overnight fast and 48 h after the last exercise bout to avoid the transient effect of exercise on transcriptional activity, which can last up to 24 h (47).
Experimental trial.
This experiment was performed at least 48 h after subjects underwent the muscle biopsy procedure to ensure that there would be no difficulties in completing the exercise task. Subjects reported to the laboratory in the morning after an overnight fast. Teflon catheters (Venflon; Becton Dickinson, Plymouth, UK) were inserted into the antecubital veins of each arm to allow stable isotope infusion and repeated blood sampling. After baseline blood and expired breath samples were obtained, ubjects received a [6,6-2H2]glucose (priming dose = 26 μmol·kg−1, continuous rate = 0.7 μmol·kg−1·min−1) and [U-13C]palmitate infusion (no priming dose, continuous rate = 16.5 nmol·kg−1·min−1) using a calibrated syringe pump (Asena GS Syringe Pump; Alaris Medical Systems, Basingstoke, UK). Subjects then rested in a semisupine position for 60 min, allowing sufficient time for isotopic equilibrium to be reached, before commencing steady-state cycling for 60 min at 70% V˙O2max. Blood and expired breath samples were collected at the end of the resting period and at 15-min intervals during exercise for the determination of isotopic enrichments, metabolite concentrations, and measures of V˙O2 and V˙CO2.
On completion of the 60-min steady-state cycle test, the ergometer was adjusted to the cadence-dependent (linear) mode, and subjects completed a set amount of work (1017 ± 73 kJ) as fast as possible. The total amount of work to be performed was calculated using the formula:
The only information available to the subjects during the time trial was elapsed work and percentage of work performed (i.e., 0% at the start and 100% on completion). No blood or respiratory measures were taken, and every effort was made to minimize any possible distractions. These are standard testing procedures used in our laboratory (7).
Training intervention.
Training consisted of nine aerobic training (AT) and nine HIT sessions spread during a 3-wk period. All training sessions took place in the laboratory under the careful supervision of the experimenters. AT was performed on an electromagnetically braked cycle ergometer (Lode Excalibur Sport) and consisted of 90-min continuous cycling at ∼70% V˙O2max. HIT was performed on the subject's own bicycle, attached to a stationary trainer, fitted with power measuring SRM cranks (Schoberer Rad Messtechnik, Julich, Germany), and consisted of a 20-min warm-up followed by 8 × 5-min all-out efforts interspersed with 1 min of recovery. The metabolic demands of this type of training have been reported elsewhere (43). This setup, using a stationary trainer and the subject's own bicycle, was chosen for the self-paced HIT sessions because it enabled the subject to instantly adjust power output through changing gear and/or pedaling speed. SRM cranks were factory-calibrated before the start of the study, and the zero offset was checked on a daily basis according to the manufacturer's instructions.
To manipulate preexercise muscle glycogen content, each group trained according to different schedules. HIGH trained once daily, alternating between AT on day 1 and HIT the following day, whereas LOW trained twice every second day, first performing AT and then, 1 h later, performing HIT. All training sessions were performed in the morning after an overnight fast, and subjects continued to fast until the entire training session was completed. Water intake was allowed during the training sessions. We did not obtain muscle biopsies to confirm muscle glycogen content; however, it was assumed that the long (20-24 h) recovery period in HIGH would allow adequate time for glycogen resynthesis between each training session (providing sufficient CHO intake), whereas the short (1 h) recovery period in LOW would ensure that HIT began with reduced muscle glycogen. This assumption is confirmed by a recent study using exactly the same training intervention and a similarly trained population of elite male cyclists (48). In that study, subjects training twice every second day began HIT with 50% less muscle glycogen than subjects training just once daily. Self-selected power output during HIT was used as a measure of training intensity.
Posttraining measures.
The nature and timing of posttraining measures were identical with those of baseline measures. A muscle biopsy sample was obtained 48 h after the final training session, using the same leg as the baseline sample but separated by 3-5 cm from the first incision. A further 48 h later, subjects performed a 60-min steady-state cycle test at the same absolute workload as baseline measures (i.e., ∼70% of pretraining V˙O2max). This was immediately followed by a time trial designed to last ∼60 min.
Diet Control
The nutritional status of subjects was controlled for 24 h before the experimental measures (muscle biopsy, 60-min steady-state cycle test, and time trial). Subjects were provided with a standard diet consisting of 67.5% CHO (8 g·kg−1 body mass), 13.5% protein, and 19% fat. Throughout the 3-wk training intervention, subjects were asked to maintain a high CHO diet and were given detailed instructions on how to achieve this.
Analyses
Blood samples (10 mL) were collected into EDTA-containing tubes (Becton Dickinson, Plymouth, UK) and centrifuged at 3000 rpm for 10 min at 4°C. Aliquots of the plasma were immediately frozen in liquid nitrogen and stored at −80°C until further analysis. Plasma glucose (Glucose HK; ABX Diagnostics, Shefford, UK) and free fatty acid (FFA; NEFA-C Kit; Alpha Laboratories, Eastleigh, Hampshire, UK) concentrations were analyzed using a semiautomatic analyzer (Cobas Mira Plus; ABX Diagnostics).
Expired breath samples were analyzed for 13C/12C ratio by gas chromatography continuous-flow isotope ratio mass spectrometry (GC-IRMS; Trace GC Ultra and IRMS Delta Plus XP, both Thermo Finnigan, Herts, UK). For the determination of plasma palmitate kinetics, FFA were extracted from plasma, isolated by solid-phase extraction, and derivatized to their methyl esters. Palmitate concentration was determined on an analytical GC equipped with flame ionization detection (FID), using heptadecanoic acid as the internal standard, and the tracer-to-tracee (TTR) ratio of [U-13C]palmitate was determined using GC combustion IRMS. After derivatization with heptafluorobutyric acid, plasma enrichment was determined by electron ionization GC-mass spectrometry (Agilent 6890N (GC) and Agilent 5973N (MS); both Agilent Technologies, Stockport, UK).
Glycogen Content
A section of muscle (∼20 mg) was freeze-dried and dissected free of connective tissue, blood, and other nonmuscle tissue. Samples were then homogenized using pestle and mortar, and 500 μL of 1 M hydrochloric acid was added to 2- to 3-mg portions and heated at 100°C for 3 h. After cooling to room temperature, samples were neutralized with 250 μL of a Tris-KOH mixture (1.44 g of Tris and 12 g of KOH in 100 mL of distilled water saturated with KCl), centrifuged at 3000 rpm for 10 min at a temperature of 4°C, and the resulting supernatant analyzed enzymatically for glucose concentration (Glucose HK; ABX Diagnostics). Muscle glycogen concentrations were expressed as millimoles of glucosyl units per kilogram of dry mass (mmol·kg−1 dm).
Protein Content
Antibodies.
The following primary antibodies were used at a 1:1000 dilution: rabbit polyclonal fatty acid translocase CD36 (FAT/CD36, ab36977; AbCam, Cambridge, MA); mouse monoclonal cytochrome c oxidase subunit 2 (COX2, A11142; Molecular Probes, Invitrogen, Carlsbad, CA); subunit 5 (COX5, A21351; Molecular Probes); chicken polyclonal antibody recognizing β-HAD (ab37673; AbCam); rabbit polyclonal antibody recognizing eukaryotic elongation factor 2 (eEF2, no. 2332; Cell Signaling, Danvers, MA); and a mouse monoclonal antibody for GLUT4 (sc-53566; Santa Cruz Biotechnology, Santa Cruz, CA).
Tissue preparation.
Muscle samples were powdered on dry ice, weighed, and aliquoted into 1.5-mL Eppendorf tubes for subsequent analysis. Approximately 30 mg of powdered tissue was added to 250 μL of lysis buffer (50 mM Tris, pH 7.5, 250 mM sucrose, 1 mM EDTA, 1 mM EGTA, 1% Triton X-100, 1 mM NaVO4, 50 mM NaF, 0.10% dithiothreitol (DTT), 0.50% protease inhibitor cocktail (PIC). The samples were mixed at 1200 rpm and 4°C for 30 min and then homogenized on ice for ∼30 s at slow speed (Polytron, Montreal, Canada), the lysate was centrifuged for 5 min at 12,000 rpm, and the supernatant was removed for protein determination. Protein concentration was determined using the DC protein assay (BioRad, Hercules, CA).
Western blot.
Equal aliquots of muscle protein were solubilized in Laemmli sample buffer (250 mM Tris-HCl, pH 6.8, 2% SDS, 10% glycerol, 0.01% bromophenol blue, 5% β-mercaptoethanol), and the samples were boiled for 5 min and separated by SDS-polyacrylamide gel electrophoresis on a 7.5% or 15% acrylamide gel. After electrophoresis, proteins were transferred to a nitrocellulose membrane (Protran; Whatman, Dassel, Germany) at 100 V for 1 h. The membrane was blocked for 1 h in 3% milk in Tris-buffered saline + 0.1% Tween (TBST). Membranes were incubated overnight at 4°C with appropriate primary antibody in TBST (1:1000 dilutions). The membrane was then washed three times in TBST before incubation for 1 h at room temperature with peroxidase-conjugated IgG secondary antibody in TBST (1:10,000 dilution; Pierce, Rockford, IL). Antibody binding was detected using enhanced chemiluminescence horse radish peroxidase (HRP) substrate detection kit (Millipore, Billerica, MA), and imaging and band quantification were carried out using a ChemiGenius Bioimaging Gel Doc System (Syngene, Cambridge, UK). All individual presamples and postsamples were run on the same gel to minimize gel-to-gel variation, whereas protein content was normalized to eEF2, which served as an internal control.
Calculations
Rates of whole-body CHO and fat oxidation were calculated using stoichiometric equations (22), with the assumption that protein oxidation was negligible:
Rates of appearance (Ra) and disappearance (Rd) of glucose and palmitate were calculated using the single-pool non-steady-state equations of Steele (39), modified for use with stable isotopes (46):
where F is the infusion rate (μmol·kg−1·min−1); V is the volume of distribution for glucose or palmitate (0.16 and 0.04 L·kg−1, respectively); C1 and C2 are the plasma glucose or palmitate concentrations (μmol·L−1) at time points t1 and t2, respectively; and E1 and E2 are plasma glucose or palmitate enrichments (TTR) at time points t1 and t2, respectively. 13CO2 production (Pr13CO2; mol·min−1) from the infused palmitate was calculated as follows:
where TTRCO2 is the breath 13C/12C ratio at a given time point, V˙CO2 is carbon dioxide production (L·min−1), k is the volume of 1 mol of CO2 (22.4 L·mol−1), and Ar is the fractional recovery of 13C label recovery in breath CO2 after infusion of labeled acetate, calculated as follows:
where F is infusion rate of [1,2-13C]acetate (mol·min−1). Plasma palmitate oxidation (Rox; mol·min−1) can then be calculated as follows:
where Rd palmitate is the rate of disappearance of palmitate (mol·min−1), F is the palmitate infusion rate (mol·min−1), and 16 is the number of carbon atoms in palmitate. Total FFA oxidation was calculated by dividing plasma oxidation by the fractional contribution of plasma palmitate to total plasma FFA concentration (plasma palmitate concentrations were, on average, 24% of total plasma FFA). Muscle-derived triacylglycerol (TG) use was estimated by subtracting plasma FFA oxidation from total fat oxidation. However, it should be mentioned that these stable isotope methods do not differentiate between muscle- and lipoprotein-derived TG uses. Nonetheless, these methods are routinely used to study exercise metabolism because the contribution of lipoprotein-derived TG oxidation to energy expenditure is assumed to be minimal. Rd glucose was taken to represent plasma glucose oxidation because 100% of Rd glucose is oxidized during exercise (21). Therefore, muscle glycogen oxidation was calculated by subtracting plasma glucose oxidation from total CHO oxidation. Calculations of tracer kinetics and substrate metabolism were performed during the final 30 min of exercise (30- to 60-min period).
Statistical Analysis
All data were analyzed using a two-factor ANOVA, with one between-subject factor (group; HIGH vs LOW) and one within-subject factor (time; pretraining vs posttraining). The level of significance was set at P < 0.05, and significant interactions were followed up with the Tukey honestly significant difference post hoc test. All data are presented as means ± SEM unless otherwise stated.
RESULTS
Performance Data
Self-selected training intensity.
HIT power output increased throughout the training period (main effect for time, P < 0.001; Fig. 1). Mean power output was 323 ± 9 W in HIGH compared with 297 ± 8 W in LOW (main effect for group, P < 0.05). This represents 87% ± 1% and 78% ± 2% Wmax for HIGH and LOW groups, respectively.
FIGURE 1: Mean power output during each HIT session performed with either HIGH or LOW muscle glycogen. Values are means ± SEM (n = 7 HIGH and n = 7 LOW). #Significant difference between HIGH and LOW groups, P < 0.05.
Time trial performance before and after training.
After training, mean power output during the time trial increased from 271 ± 13 to 298 ± 13 W in HIGH and from 278 ± 11 to 307 ± 10 W in LOW (main effect for time, P < 0.001; Fig. 2). Time taken to complete the task decreased from 62.10 ± 1.49 to 56.37 ± 1.17 min in HIGH and from 61.90 ± 1.12 to 56.12 ± 1.22 min in LOW (main effect for time, P < 0.001). Performance was enhanced by a similar extent in both groups (10.2% in HIGH and 10.5% in LOW).
FIGURE 2: Time trial performance before and after training with HIGH or LOW muscle glycogen. Values are means ± SEM (n = 7 HIGH and n = 7 LOW). *Significant difference between pretraining and posttraining, P < 0.05.
Cycling at 70% V˙O2max before and after Training
V˙O2, RER, and whole-body substrate metabolism.
Subjects cycled at the same absolute workload before and after training (241 ± 13 W in HIGH and 247 ± 16 W in LOW). V˙O2 was unaffected by training (P > 0.05; Table 2). RER was unaffected by training in HIGH but decreased after training in LOW (group × time interaction, P < 0.001; Table 2). Accordingly, there was a significant decrease in the rate of CHO oxidation (from 220 ± 8 to 194 ± 10 μmol·kg−1·min−1) and a significant increase in the rate of fat oxidation (from 26 ± 2 to 34 ± 2 μmol·kg−1·min−1) after training in LOW (P < 0.01; Table 2).
TABLE 2: Oxygen consumption, RER, and substrate oxidation during steady-state cycling at 70% V˙O2max (30- to 60-min period) before and after training with HIGH or LOW muscle glycogen.
Tracer data.
Plasma glucose and palmitate kinetics are shown in Table 3, and rates of substrate oxidation are shown in Table 2. Ra and Rd glucose were unaffected by training in HIGH but decreased after training in LOW (group × time interaction, P < 0.05; Table 3). Because Rd glucose represents plasma glucose oxidation, there was a significant decrease in plasma glucose oxidation (from 39 ± 3 to 33 ± 2 μmol·kg−1·min−1) after training in LOW (P < 0.05; Table 2). Likewise, tracer-estimated rates of muscle glycogen oxidation decreased significantly (from 181 ± 8 to 161 ± 9 μmol·kg−1·min−1) after training in LOW (P < 0.05; Table 2). Ra and Rd palmitate increased after training (main effect for time, P < 0.05; Table 3); however, because Rox palmitate remained unchanged, %Rdox decreased after training (main effect for time, P < 0.05; Table 3). Estimated rates of plasma FFA oxidation were unaffected by training (Table 2). Oxidation of muscle-derived TG increased (from 16 ± 1 to 23 ± 1 μmol·kg−1·min−1) after training in LOW only (group × time interaction, P < 0.05; Table 2).
TABLE 3: Glucose and palmitate kinetics during steady-state cycling at 70% V˙O2max (30- to 60-min period) before and after training with HIGH or LOW muscle glycogen.
Relative contribution of substrates to energy expenditure.
The relative contribution of fat to energy expenditure during exercise was unaffected by training in HIGH but increased significantly (from 32% ± 2% to 40% ± 2%) after training in LOW (group × time interaction, P < 0.05; Fig. 3). This increase was due to higher rates of muscle-derived TG oxidation (Table 2), which provided 20% ± 1% of energy expenditure before training and 28% ± 2% of energy expenditure after training (P < 0.05; Fig. 3). The relative contribution of muscle glycogen to energy expenditure during exercise was unaffected by training in HIGH but decreased (from 57% ± 2% to 50% ± 2%) after training in LOW (group × time interaction, P < 0.05; Fig. 3). In LOW, there was also a small, nonsignificant decrease in the contribution of plasma glucose (from 12% ± 1% to 10% ± 1%).
FIGURE 3: Relative contribution of substrates to energy expenditure, during the 30- to 60-min exercise period, before and after training with HIGH or LOW muscle glycogen. Values are means ± SEM (n = 7 HIGH and n = 7 LOW). *Significant difference between pretraining and posttraining, P < 0.05.
Skeletal Muscle Adaptations
Muscle glycogen content.
Resting muscle glycogen content increased after training by 18% and 36% in HIGH and LOW, respectively, with no difference between groups (main effect for time, P < 0.001; Fig. 4).
FIGURE 4: Resting muscle glycogen content before and after training with HIGH or LOW muscle glycogen. Values are means ± SEM (n = 7 HIGH and n = 7 LOW). *Significant difference between pretraining and posttraining, P < 0.05.
Protein content.
COX2 and COX5 protein content was unaffected by training (P > 0.05; data not shown). FAT/CD36 protein content increased after training (main effect for time, P < 0.05; Fig. 5B) and tended to be more pronounced after training in LOW than in HIGH (41.4% vs 11.5%), but this was not significant. β-HAD protein content increased by 43% after training in LOW but decreased by 20% after training in HIGH (group × time interaction, P < 0.01; Fig. 5C). In contrast to the beneficial effect on proteins involved in fatty acid uptake and oxidation, the glucose transporter GLUT4 tended not to increase as much in the LOW group (LOW = 7.1% ± 8.4% vs HIGH = 20.6% ± 9.7%; Fig. 5A).
FIGURE 5: Effect of training with HIGH or LOW muscle glycogen on resting protein content for GLUT4 (A), FAT/CD36 (B), and β-HAD (C). Values are expressed as percentage change relative to pretraining levels. Data shown are expressed as means ± SEM (n = 7 HIGH and n= 7 LOW). *Significant difference between pretraining and posttraining, P < 0.05. #Significant difference between HIGH and LOW groups, P < 0.05.
DISCUSSION
This study determined the effects of training with low muscle glycogen on exercise performance, substrate metabolism, and molecular adaptations in skeletal muscle. To manipulate muscle glycogen content each group trained according to different schedules. HIGH trained once daily, alternating between AT on day 1 and HIT the following day, whereas LOW trained twice every second day, first performing AT and then, 1 h later, performing HIT. We therefore assume that HIGH began all training sessions with normal glycogen stores, whereas LOW commenced HIT with reduced muscle glycogen (see Methods for further details).
One of the main findings of the present study was that whole-body fat oxidation, during a moderate-intensity exercise, was increased after training with low muscle glycogen (Table 2). Similar findings have been reported recently elsewhere (49). In that study, however, measures of substrate metabolism were limited to indirect calorimetry, whereas we also applied stable isotope tracers for a more detailed investigation of changes in substrate metabolism. By infusing [U-13C]palmitate, we were able to determine plasma palmitate kinetics, plasma FFA oxidation, and muscle-derived TG oxidation. We found that plasma FFA oxidation was similar before and after training in both groups but that muscle-derived TG oxidation was significantly increased after training with low muscle glycogen (Table 2 and Fig. 3). There has been considerable debate regarding the role of muscle TG as a substrate during exercise (for reviews, see Kiens [26] and Watt et al. [45]). However, using three different methodologies (stable isotopes, biochemical extraction, and immunofluorescence microscopy), Stellingwerff et al. (41) clearly demonstrated significant muscle TG use during prolonged submaximal exercise. Endurance training has also been shown to increase muscle TG use during exercise in previously untrained men (32). In the present study, training with high muscle glycogen did not increase TG utilization during exercise, suggesting that several years of endurance training had already maximized the capacity for muscle TG use in these subjects. Therefore, it is interesting to find that training for just 3 wk with low muscle glycogen can significantly increase muscle TG use in a group of similarly well-trained subjects. There are several possible mechanisms for increased muscle TG oxidation during exercise, including elevated preexercise muscle TG concentrations, increased hormone-sensitive lipase activity, and/or increased hormone-sensitive lipase migration to lipid droplets.
A [6,6-2H2]glucose tracer was also infused to estimate plasma glucose and muscle glycogen oxidation. In the present study, we have shown that the reduction in whole-body CHO oxidation during exercise, after training with low muscle glycogen, was due to the reduced rates of muscle glycogenolysis because the decrease in plasma glucose oxidation was only modest when compared with the decrease in muscle glycogen oxidation (Table 2 and Fig. 3). This reduction in muscle glycogenolysis occurred despite higher resting muscle glycogen levels (Fig. 4). Because substrate availability cannot explain the changes in CHO metabolism, we suggest that training with low muscle glycogen evokes chronic adaptations within the skeletal muscle that causes a down-regulation in muscle glycogenolysis. A likely candidate for this down-regulation is the pyruvate dehydrogenase (PDH) complex. PDH catalyzes the conversion of pyruvate to acetyl-CoA and thus regulates the entry of CHO substrates into the mitochondria (34). Acute studies have shown that commencing endurance exercise with low muscle glycogen increases mRNA and protein content of PDH kinase 4 (28,33), an enzyme that phosphorylates and inactivates PDH. Training with reduced CHO availability may cause persistent decreases in PDH activity that reduces muscle glycogenolysis during exercise, as has been shown previously after high-fat feeding in humans (42).
Changes in substrate selection could be the result of an increase in the capacity to oxidize fat, a decreased capacity to oxidize CHO, or a combination of both. The current study suggests the latter possibility. Fat oxidation during exercise can be regulated at several sites, including adipose tissue lipolysis, delivery of fatty acids to the muscle, transport across the plasma membrane, lipolysis of muscle TG stores, transport within the cytosol, and transport into the mitochondria (for reviews, see Kiens [26] and Spriet [38]). Fatty acid transport is a highly regulated process involving several specialized proteins (4,19,31). Unfortunately, we were unable to measure most of these proteins. Nonetheless, we present data for FAT/CD36, which is known to play an important role in fatty acid transport across both the plasma and mitochondrial membranes (19) and contributes to the regulation of mitochondrial fatty acid oxidation (18). In the present study, FAT/CD36 increased with training (Fig. 5B), and although not statistically significant, this increase tended to be more pronounced after training with low muscle glycogen (41.4% vs 11.5%), which may explain higher rates of fat oxidation during exercise.
Fat oxidation is also regulated by the activity of mitochondrial enzymes such as β-HAD (25). We observed a 43% increase in β-HAD protein content after training with low muscle glycogen (Fig. 5C), which is consistent with the 27.7% increase in β-HAD activity measured by Yeo et al. (49). Studies have shown that β-HAD is highly sensitive to changes in substrate availability. For example, Helge and Kiens (16) reported a 25% increase in β-HAD activity after 7 wk of adaptation to a high-fat diet (62% energy from fat). This increase occurred irrespective of training and demonstrates that altered substrate availability (e.g., increased dietary fat) influences β-HAD activity. Although no clear molecular mechanism has been identified, it is possible that the flux through β-oxidation plays an important role in the regulation of β-HAD content and activity. We did not alter macronutrient intake; however, performing exercise in a glycogen-depleted state increases circulating free fatty acids (40), and increasing the availability of free fatty acids is known to increase their oxidation during a high-intensity exercise (37). Therefore, it is reasonable to assume that rates of fat oxidation would have been elevated to a greater extent during training sessions performed with low muscle glycogen and that this may have contributed to the changes in β-HAD content that we observed.
This study was the first to look at the effect of training with low muscle glycogen on the glucose transporter GLUT4. GLUT4 is the rate-limiting enzyme in muscle glucose utilization (29) and GLUT4 levels increase with training in animals (35) and both disease-free and type 2 diabetic humans (8). Furthermore, GLUT4 is significantly higher in endurance athletes than sedentary controls (13). Consistent with these results, GLUT4 levels tended to increase in the HIGH group (20.6% ± 9.7%). However, the change in GLUT4 levels tended to be smaller in the LOW group (7.1% ± 8.4%). Together with the β-HAD and FAT/CD36 data, this suggests that the capacity for fatty acid oxidation increased more in the LOW group and the capacity for CHO oxidation increased less, resulting in the changes in substrate metabolism that we observed during steady-state exercise. This also suggests that training with low muscle glycogen may be counterproductive for athletes who compete in high-intensity events where CHO oxidation plays a significant role in performance and that this type of training may be more suited to preparation for ultraendurance activities.
Another major finding of the present study is that despite reduced power outputs during HIT (Fig. 1), training with low muscle glycogen enhanced time trial cycling performance by a similar amount to that of training with high muscle glycogen (Fig. 2). To date, only two previous studies have investigated the effect of training with low muscle glycogen (12,49). Hansen et al. (12) had untrained subjects perform single-leg knee extensor exercise (1 h at 75% maximal power output) for 10 wk using a similar twice-every-second-day (LOW) or once-daily (HIGH) design. Time to fatigue at 90% maximal power output increased after training in both legs; however, time to fatigue was greater in the LOW leg compared with the HIGH leg (19.7 ± 2.4 vs 11.9 ± 1.3 min). In that study, untrained subjects performed a fixed (submaximal) amount of work in each training session, although higher glycogen stores would normally allow for exercise at higher intensities and/or longer durations. In the present study, well-trained cyclists worked at self-selected intensities, just as they would in their normal training sessions. As expected, training with low muscle glycogen resulted in lower power outputs in HIT sessions. The finding that time trial performance was still improved by a similar extent in both groups after the 3-wk training period is in direct agreement with the recent work of Yeo et al. (49) and suggests that the amount of work performed during training may not be critical. In this regard, we suggest that the additional "stress" of training with low glycogen compensates for a slight reduction in physical performance during training. Nonetheless, under the specific test conditions of the present study, training with low muscle glycogen did not translate into better performance and does not seem to offer a worthwhile benefit for already well-trained athletes. Whether the increased capacity for fat oxidation translates into better performance during a longer-duration test of exercise performance or after a longer period of adaptation remains to be seen.
An important practical question is whether additional strategies could be used to further maximize the muscle "signaling state" during exercise with low glycogen. Glycogen is known to affect the activity of several important signaling proteins (30), but whether the amount of glycogen or the localization of glycogen particles is the important factor in mediating this response is unknown. Further, it may be that nutritional interventions in a low glycogen state could either enhance the adaptive response or reduce the drop-off in training intensity without affecting signaling. With regard to the latter, one technique that may allow for greater exercise intensity during low glycogen training is the use of a CHO mouth rinse. We have previously shown that simply rinsing the mouth with a 6% CHO solution was sufficient to improve performance in a 60-min time trial by ∼3% (5), approximately half of the difference seen between the HIGH and LOW training groups in the current study. Importantly, Chambers et al. (6) have recently shown that the benefit of the CHO mouth rinse is due to brain receptor modulation, and as a result, this intervention would not be expected to alter the metabolic state of skeletal muscle.
In conclusion, training with low muscle glycogen reduced self-selected training intensity and, in terms of performance, was no more effective than training with high muscle glycogen. However, training with low muscle glycogen increased muscle TG oxidation during a moderate-intensity exercise, which may have been due to enhanced metabolic adaptations in skeletal muscle. Future studies should examine the signaling pathways and regulatory mechanisms that govern metabolic adaptations to endurance training with low muscle glycogen.
This study was supported by a research grant of GlaxoSmithKline, Nutritional Healthcare, R&D. Work in the Functional Molecular Biology Laboratory was supported by a grant from the Wellcome Trust (077426). The results of the present study do not constitute endorsement by the American College of Sports Medicine.
REFERENCES
1. Achten J, Halson SL, Moseley L, Rayson MP, Casey A, Jeukendrup AE. Higher dietary carbohydrate content during intensified running training results in better maintenance of performance and mood state.
J Appl Physiol. 2004;96(4):1331-40.
2. Bergstrom J. Percutaneous needle biopsy of skeletal muscle in physiological and clinical research.
Scand J Clin Lab Invest. 1975;35(7):609-16.
3. Bergström J, Hermansen L, Hultman E, Saltin B. Diet, muscle glycogen and physical performance.
Acta Physiol Scand. 1967;71:140-50.
4. Bonen A, Miskovic D, Kiens B. Fatty acid transporters (FABPpm, FAT, FATP) in human muscle.
Can J Appl Physiol. 1999;24(6):515-23.
5. Carter JM, Jeukendrup AE, Jones DA. The effect of carbohydrate mouth rinse on 1-h cycle time trial performance.
Med Sci Sports Exerc. 2004;36(12):2107-11.
6. Chambers E, Bridge MW, Jones DA. Carbohydrate sensing in the human mouth: effects on
exercise performance and brain activity.
J Physiol. 2009;587(8):1779-94.
7. Currell K, Jeukendrup AE. Validity, reliability and sensitivity of measures of sporting performance.
Sports Med. 2008;38(4):297-316.
8. Dela F, Ploug T, Handberg A, et al. Physical training increases muscle GLUT4 protein and mRNA in patients with NIDDM.
Diabetes. 1994;43(7):862-5.
9. Evans WJ, Phinney SD, Young VR. Suction applied to a muscle biopsy maximizes sample size.
Med Sci Sports Exerc. 1982;14(1):101-2.
10. Febbraio MA, Steensberg A, Walsh R, et al. Reduced glycogen availability is associated with an elevation in HSP72 in contracting human skeletal muscle.
J Physiol. 2002;538(Pt 3):911-7.
11. Gollnick PD, Riedy M, Quintinskie JJ, Bertocci LA. Differences in metabolic potential of skeletal muscle fibres and their significance for metabolic control.
J Exp Biol. 1985;115:191-9.
12. Hansen AK, Fischer CP, Plomgaard P, Andersen JL, Saltin B, Pedersen BK. Skeletal muscle adaptation: training twice every second day vs. training once daily.
J Appl Physiol. 2005;98(1):93-9.
13. Hardin D, Azzarelli B, Edwards J, et al. Mechanisms of enhanced insulin sensitivity in endurance-trained athletes: effects on blood flow and differential expression of GLUT4 in skeletal muscle.
J Clin Endocrinol Metab. 1995;80(8):2437-46.
14. Hawley JA, Palmer GS, Noakes TD. Effects of 3 days of carbohydrate supplementation on muscle glycogen content and utilisation during a 1-h cycling performance.
Eur J Appl Physiol Occup Physiol. 1997;75(5):407-12.
15. Hawley JA, Tipton KD, Millard-Stafford ML. Promoting training adaptations through nutritional interventions.
J Sports Sci. 2006;24(7):709-21.
16. Helge JW, Kiens B. Muscle enzyme activity in humans: role of substrate availability and training.
Am J Physiol. 1997;272(5 Pt 2):R1620-4.
17. Henriksson J. Training induced adaptation of skeletal muscle and metabolism during submaximal exercise.
J Physiol. 1977;270(3):661-75.
18. Holloway GP, Jain SS, Bezaire V, et al. FAT/CD36-null mice reveal that mitochondrial FAT/CD36 is required to upregulate mitochondrial fatty acid oxidation in contracting muscle.
Am J Physiol Regul Integr Comp Physiol. 2009;297:R960-7.
19. Holloway GP, Luiken JJ, Glatz JF, Spriet LL, Bonen A. Contribution of FAT/CD36 to the regulation of skeletal muscle fatty acid oxidation: an overview.
Acta Physiol (Oxf). 2008;194(4):293-309.
20. Irrcher I, Adhihetty PJ, Joseph AM, Ljubicic V, Hood DA. Regulation of mitochondrial biogenesis in muscle by endurance exercise.
Sports Med. 2003;33(11):783-93.
21. Jeukendrup AE, Raben A, Gijsen A, et al. Glucose kinetics during prolonged exercise in highly trained human subjects: effect of glucose ingestion.
J Physiol. 1999;515(Pt 2):579-89.
22. Jeukendrup AE, Wallis GA. Measurement of substrate oxidation during exercise by means of gas exchange measurements.
Int J Sports Med. 2005;26(suppl 1):S28-37.
23. Karlsson J, Saltin B. Diet, muscle glycogen, and endurance performance.
J Appl Physiol. 1971;31(2):203-6.
24. Keller C, Steensberg A, Pilegaard H, et al. Transcriptional activation of the IL-6 gene in human contracting skeletal muscle: influence of muscle glycogen content.
FASEB J. 2001;15(14):2748-50.
25. Kiens B. Effect of endurance training on fatty acid metabolism: local adaptations.
Med Sci Sports Exerc. 1997;29(5):640-5.
26. Kiens B. Skeletal muscle lipid metabolism in exercise and insulin resistance.
Physiol Rev. 2006;86:205-43.
27. Kiens B, Essen-Gustavsson B, Christensen NJ, Saltin B. Skeletal muscle substrate utilization during submaximal exercise in man: effect of endurance training.
J Physiol. 1993;469:459-78.
28. Kiilerich K, Gudmundsson M, Birk JB, et al. Low muscle glycogen and elevated plasma free fatty acid modify but do not prevent exercise-induced PDH activation in human skeletal muscle.
Diabetes. 2010;59(1):26-32.
29. Lombardi A, Moller D, Loizeau M, Girard J, Leturque A. Phenotype of transgenic mice overexpressing GLUT4 and hexokinase II in muscle.
FASEB J. 1997;11(13):1137-44.
30. McBride A, Hardie DG. AMP-activated protein kinase-a sensor of glycogen as well as AMP and ATP?
Acta Physiol (Oxf). 2009;196(1):99-113.
31. Nickerson JG, Alkhatteb H, Benton CR, et al. Greater transport efficiencies of the membrane fatty acid transporters FAT/CD36 and FATP4 compared with FABPpm and FATP1 and differential effects on the fatty acid esterification and oxidation in rat skeletal muscle.
J Biol Chem. 2009;284(4):16522-30.
32. Phillips SM, Green HJ, Tarnopolsky MA, Heigenhauser GF, Hill RE, Grant SM. Effects of training duration on substrate turnover and oxidation during exercise.
J Appl Physiol. 1996;81(5):2182-91.
33. Pilegaard H, Keller C, Steensberg A, et al. Influence of pre-exercise muscle glycogen content on exercise-induced transcriptional regulation of metabolic genes.
J Physiol. 2002;541(Pt 1):261-71.
34. Pilegaard H, Neufer PD. Transcriptional regulation of pyruvate dehydrogenase kinase 4 in skeletal muscle during and after exercise.
Proc Nutr Soc. 2004;63(2):221-6.
35. Rodnick K, Holloszy JO, Mondon C, James D. Effects of exercise training on insulin-regulatable glucose-transporter protein levels in rat skeletal muscle.
Diabetes. 1990;39(11):1425-9.
36. Romijn JA, Coyle EF, Sidossis LS, et al. Regulation of endogenous fat and carbohydrate metabolism in relation to exercise intensity and duration.
Am J Physiol. 1993;265(3 Pt 1):E380-91.
37. Romijn JA, Coyle EF, Sidossis LS, Zhang XJ, Wolfe RR. Relationship between fatty acid delivery and fatty acid oxidation during strenuous exercise.
J Appl Physiol. 1995;79(6):1939-45.
38. Spriet LL. Regulation of skeletal muscle fat oxidation during exercise.
Med Sci Sports Exerc. 2002;34(9):1477-84.
39. Steele R. Influences of glucose loading and of injected insulin on hepatic glucose output.
Ann N Y Acad Sci. 1959;82:420-30.
40. Steensberg A, van Hall G, Keller C, et al. Muscle glycogen content and glucose uptake during exercise in humans: influence of prior exercise and dietary manipulation.
J Physiol. 2002;541(Pt 1):273-81.
41. Stellingwerff T, Boon H, Jonkers RAM, et al. Significant intramyocellular lipid use during prolonged cycling in endurance-trained males as assessed by three different methodologies.
Am J Physiol Endocrinol Metab. 2007;293:E1715-23.
42. Stellingwerff T, Spriet LL, Watt MJ, et al. Decreased PDH activation and glycogenolysis during exercise following fat adaptation with carbohydrate restoration.
Am J Physiol Endocrinol Metab. 2006;290(2):E380-8.
43. Stepto NK, Martin DT, Fallon KE, Hawley JA. Metabolic demands of intense aerobic interval training in competitive cyclists.
Med Sci Sports Exerc. 2001;33(2):303-10.
44. van Loon LJ, Greenhaff PL, Constantin-Teodosiu D, Saris WH, Wagenmakers AJ. The effects of increasing exercise intensity on muscle fuel utilisation in humans.
J Physiol. 2001;536(Pt 1):295-304.
45. Watt MJ, Heigenhauser GF, Spriet LL. Intramuscular triacylglycerol utilization in human skeletal muscle during exercise: is there a controversy?
J Appl Physiol. 2002;93:1185-95.
46. Wolfe RR, Chinkes DL.
Isotope Tracers in Metabolic Research: Principles and Practice of Kinetic Analysis. 2nd ed. Hoboken (NJ): John Wiley and Sons; 2005. p. 38-42.
47. Yang Y, Creer A, Jemiolo B, Trappe S. Time course of myogenic and metabolic gene expression in response to acute exercise in human skeletal muscle.
J Appl Physiol. 2005;98(5):1745-52.
48. Yeo WK, McGee SL, Carey AL, et al. Acute signalling responses to intense endurance training commenced with low or normal muscle glycogen.
Exp Physiol. 2009;95(2):351-8.
49. Yeo WK, Paton CD, Garnham AP, Burke LM, Carey A, Hawley JA. Skeletal muscle adaptation and performance responses to once a day versus twice every second day endurance training regimens.
J Appl Physiol. 2008;105(5):1519-26.
